Image capture for spectral profiling of objects in a scene

摘要

Image data of a scene is captured. Spectral profile information is obtained for the scene. A database of plural spectral profiles is accessed, each of which maps a material to a corresponding spectral profile reflected therefrom. The spectral profile information for the scene is matched against the database, and materials for objects in the scene are identified by using matches between the spectral profile information for the scene against the database. Metadata which identifies materials for objects in the scene is constructed, and the metadata is embedded with the image data for the scene.

主权项

1. An image capture method comprising:
capturing image data of a scene, wherein the image data is comprised of tri-stimulus device independent image data; obtaining spectral profile information for the scene; accessing a database of plural spectral profiles each of which maps a material to a corresponding spectral profile reflected therefrom; matching the spectral profile information for the scene against the database; identifying materials for objects in the scene by using matches between the spectral profile information for the scene against the database; constructing metadata which identifies materials for objects in the scene; embedding the metadata with the image data for the scene; and using the metadata to differentiate materials in the scene during post-capture rendering; and rendering the image data for the scene by using the metadata that identifies the material for objects in the scene, wherein image data having similar tri-stimulus values is rendered differently in dependence on the metadata.